For many camera trap studies it is a problem that it is not possible to distinguish individuals, therefore consecutive visits may be the same individual. Either an individual triggers a burst of captures during one visit, or the individual moves consecutively in an out of the camera view during a short time period.
To minimize double counting many studies are using a time gap (hit window), which is the length of time used to group consecutive images/videos together as single detections (given that the data are from the same camera and featuring the same species). This will vary between study designs, for example bait site will cause individuals to stay within the detections zone for a longer period compared to non-baited sites. Further, some species will be more likely to stay at the site than others.
Is there a good procedure to estimate an optimal time gap (hit window) for an existing data set?
Best regards,
Ronny
This depends mostly on the methodology (baited/non-baited; random samples/non random) and especially the analysis method.
For most of the more recent density estimate methods very short (2s etc) or no time delay have been used.
If you have enough saving capacity and battery life (which has improved in newer models) I would say you should go for the shortest interval possible. If required you can remove/filter multiple triggers in analysis. Otherwise you may miss critical data during the time interval where the camera stays innactive. And I personally prefer video captures instead of snapshots (which helps better identification).
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology